Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2017-02-21 Thread Ilia Mirkin
On Wed, Feb 22, 2017 at 12:07 AM, Jason Ekstrand wrote: > Hey Look. I'm actually reading your patch now! > > I read through the whole thing and over-all I think it looks fairly good. > > On Sun, Nov 27, 2016 at 11:23 AM, Ilia Mirkin wrote: >> >> The

Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2017-02-21 Thread Jason Ekstrand
Hey Look. I'm actually reading your patch now! I read through the whole thing and over-all I think it looks fairly good. On Sun, Nov 27, 2016 at 11:23 AM, Ilia Mirkin wrote: > The strategy is to just keep n anv_query_pool_slot entries per query > instead of one. The

Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2017-02-21 Thread Ilia Mirkin
Looks like after nearly 3 months of no reviews that lead to R-b's[1], this patch no longer applies to master. I'm abandoning it. If anyone's interested, feel free to pick it up and make it your own. Cheers, -ilia [1] Robert did look at it, but didn't have significant feedback or say what I

Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2017-02-16 Thread Robert Bragg
On Wed, Feb 15, 2017 at 11:04 PM, Ilia Mirkin wrote: > On Tue, Jan 24, 2017 at 5:27 PM, Robert Bragg wrote: >> Depending on how strictly we consider that the queries should only >> measure >> the commands they bracket then I think some

Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2017-02-15 Thread Ilia Mirkin
On Tue, Jan 24, 2017 at 5:27 PM, Robert Bragg wrote: > Depending on how strictly we consider that the queries should only measure > the commands they bracket then I think some stalling will be necessary to > serialize the work associated with a query and defer

Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2017-01-24 Thread Robert Bragg
On Tue, Jan 24, 2017 at 2:37 PM, Ilia Mirkin wrote: > On Tue, Jan 24, 2017 at 5:27 PM, Robert Bragg wrote: >>> +/* >>> + * GPR0 = GPR0 >> 2; >>> + * >>> + * Note that the upper 30 bits of GPR are lost! >>> + */ >>> +static void

Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2017-01-24 Thread Ilia Mirkin
On Tue, Jan 24, 2017 at 5:27 PM, Robert Bragg wrote: >> +/* >> + * GPR0 = GPR0 >> 2; >> + * >> + * Note that the upper 30 bits of GPR are lost! >> + */ >> +static void >> +shr_gpr0_by_2_bits(struct anv_batch *batch) >> +{ >> +

Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2017-01-24 Thread Robert Bragg
Sorry for the delay responding here; some comments below... On Tue, Jan 24, 2017 at 11:48 AM, Ilia Mirkin wrote: > 2-month ping. [ok, it hasn't been 2 months on the dot, but ... close.] > > On Tue, Jan 10, 2017 at 5:49 PM, Ilia Mirkin wrote: >> ping.

Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2017-01-24 Thread Ilia Mirkin
2-month ping. [ok, it hasn't been 2 months on the dot, but ... close.] On Tue, Jan 10, 2017 at 5:49 PM, Ilia Mirkin wrote: > ping. > > On Thu, Dec 22, 2016 at 11:14 AM, Ilia Mirkin wrote: >> Ping? Any further comments/feedback/reviews? >> >> >> On Dec

Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2017-01-10 Thread Ilia Mirkin
ping. On Thu, Dec 22, 2016 at 11:14 AM, Ilia Mirkin wrote: > Ping? Any further comments/feedback/reviews? > > > On Dec 5, 2016 11:22 AM, "Ilia Mirkin" wrote: > > On Mon, Dec 5, 2016 at 11:11 AM, Robert Bragg wrote: >> >> >> On

Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2016-12-22 Thread Ilia Mirkin
Ping? Any further comments/feedback/reviews? On Dec 5, 2016 11:22 AM, "Ilia Mirkin" wrote: On Mon, Dec 5, 2016 at 11:11 AM, Robert Bragg wrote: > > > On Sun, Nov 27, 2016 at 7:23 PM, Ilia Mirkin wrote: >> >> The strategy is to

Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2016-12-05 Thread Ilia Mirkin
On Mon, Dec 5, 2016 at 11:11 AM, Robert Bragg wrote: > > > On Sun, Nov 27, 2016 at 7:23 PM, Ilia Mirkin wrote: >> >> The strategy is to just keep n anv_query_pool_slot entries per query >> instead of one. The available bit is only valid in the last

Re: [Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2016-12-05 Thread Robert Bragg
On Sun, Nov 27, 2016 at 7:23 PM, Ilia Mirkin wrote: > The strategy is to just keep n anv_query_pool_slot entries per query > instead of one. The available bit is only valid in the last one. > > Signed-off-by: Ilia Mirkin > --- > > I think this is in a

[Mesa-dev] [PATCH] anv: implement pipeline statistics queries

2016-11-27 Thread Ilia Mirkin
The strategy is to just keep n anv_query_pool_slot entries per query instead of one. The available bit is only valid in the last one. Signed-off-by: Ilia Mirkin --- I think this is in a pretty good state now. I've tested both the direct and buffer paths with a hacked up